Height of Rectal Cancer: A Comparison between Rectoscopic and Different MRI Measurements

Abstract

Purpose. To compare rigid rectoscopy with three different MRI measurement techniques for rectal cancer height determination, all starting at the anal verge, in order to evaluate whether MRI measurements starting from the anal verge could be an alternative to rigid rectoscopy. Moreover, potential cut-off values for MRI in categorizing tumor height measurements were evaluated. Methods. In this retrospective study, 106 patients (75 men, 31 female, mean age 64±11.59 years) with primary rectal cancer underwent rigid rectoscopy as well as MR imaging. Three different measurements (MRI1–3) in T2w sagittal scans were used to evaluate the exact distance from the anal verge (AV) to the distal ending of the tumor (MRI1: two unbowed lines, AV to the upper ending of the anal canal and upper ending of the anal canal to the lower border of the tumor; MRI2: one straight line from the AV to the lower boarder of the tumor; MRI3: a curved line beginning at the AV and following the course of the rectum wall ending at the lower border of the tumor). Furthermore, agreement between the gold standard rigid rectoscopy (UICC classification: low part, 0-6 cm; mid part, 6-12 cm; and high part, >12 cm) and each MRI measuring technique was analyzed. Results. Only a fair correlation in terms of individual measures between rectoscopy and all 3 MRI measurement techniques was shown. The proposed new cut-off values utilizing ROC analysis for the three different MRI beginning at the anal verge were low 0-7.7 cm, mid 7.7-13.3 cm, and high>13.3 cm (MRI1); low 0-7.4 cm, mid 7.4-11.2 cm, and high>11.2 cm (MRI2); and low 0-7.1 cm, mid 7.1-13.7 cm, and high>13.7 cm (MRI3). For MRI1 and MRI3, the agreement to the gold standard was substantial (r=0.66, r=0.67, respectively). Conclusion. This study illustrates that MRI1 and MRI3 measures can be interchangeably used as a valid method to determine tumor height compared to the gold standard rigid rectoscopy.